Write to us at startingpoint@globe.com. To subscribe, sign up here.
The breakthrough seemed, in the words of one patient, almost “too good to be true.” But when a charming young oncologist with a sterling résumé told cancer patients that he’d developed a cutting-edge tool that could help them, many were willing to take a chance.
The doctor, Anil Potti, purportedly created algorithms that could analyze tumors and pick the best chemotherapy cocktail to treat them — effectively taking much of the guesswork out of oncology. The work of Potti and his mentor Joseph Nevins, both academic stars at Duke University Medical Center, led to clinical trials that started enrolling cancer patients in 2007.
Then, as my colleague Mike Damiano reports for the Globe Magazine, everything unraveled.
Potti had been manipulating data, a later federal investigation found; in truth, his algorithm was no better than a coin flip. And some leaders at Duke — including Sally Kornbluth, now president of the Massachusetts Institute of Technology — were slow to shut down the clinical trials even after other researchers raised serious concerns. Patients and their loved ones sued, and Duke settled in 2015 for an undisclosed sum.
The basics of the scandal have been known for years, but Mike’s reporting uncovered new details. I spoke with him about how his story came together.
Ian: Your story involves events that happened nearly two decades ago. Why write it now?
Mike: Part of the attraction was just how fully it can finally be told. I got a tip to look into Kornbluth’s role, went to a North Carolina courthouse, and found thousands of pages of records that had been collecting dust for a decade. There are emails. There’s sworn testimony where Duke leaders explain their thinking at every juncture, including Kornbluth’s detailed deposition that took place over two days. It’s rare to get to look inside a research misconduct scandal with this level of detail.
Other researchers raised concerns about Potti and Nevins’s work to Kornbluth years before Duke shut down their clinical trials. But it can be difficult to see through someone who’s trying to deceive you. How do you think about her role?
The story invites readers to grapple with what they might’ve done if they’d encountered those warning signs. Kornbluth says the warnings didn’t always point in the same direction and that she trusted her researchers. In a statement, she said that Duke’s internal review of the researchers’ work was “slow and deliberative,” but that it ultimately got to the right place.
Her critics see it differently. They don’t find it plausible that you could see all these warnings and not understand what they meant. There were people inside Duke and in the cancer research world who were extremely disillusioned.
Did Potti’s fraud harm patients?
To enroll, lung cancer patients had to undergo a biopsy. It’s an invasive procedure where a needle extracts tissue, and one patient was harmed by the biopsy. There wasn’t anything wrong with the chemotherapy drugs administered in the trials, and it’s true that some tumors are sensitive to different drugs. The claim that was bogus was that Potti’s algorithm could reliably determine which drug would work best. Some patients spent their last best chance to extend their life on a false bill of goods.
Your story mentions other recent cases of academic fraud. Are there more guardrails today?
There’s more focus on what’s referred to as the “reproducibility crisis.” The idea is that if a scientific finding is solid, it should be possible for other researchers to replicate it, and a lot of research isn’t replicable. Today there’s definitely more detection of fraud. But a lot of it is done by hobbyists online, not through formal procedures like peer review.
The Trump administration has pointed to academic fraud to bolster its argument against universities. Do they have a point?
Data falsification happens, and academia has never managed to stamp it out. Incentives reward breakthroughs like the fake one in this story. Accountability systems aren’t good at dealing with deliberate fraud. When one of these scandals happens, and they keep happening, it’s easier to diminish the credibility of scientific work.
But the administration has pointed to episodes like this to justify a crackdown on science, and another view is that, actually, science works pretty damn well. It moves toward the truth in a messy, sometimes nonlinear, fashion. And yes, when you’re working on the frontiers of knowledge, there will be mistakes, misconduct, and embarrassments. But ultimately science is self-correcting. And even this story demonstrates that.
Read Mike’s full story here.
🧩 4 Down: Financial backer | 🌤️ 92° Heat, humidity, thunderstorms
Second life: From Roxbury to Cambridge, dwindling congregations are turning their mostly empty houses of worship into low-income housing.
On your dime: The Suffolk County Sheriff’s Department spent tens of thousands on travel and food during the last fiscal year, including junkets to Vegas and the Bahamas. Federal prosecutors accused the office’s top official of extortion last week.
Charged: Authorities arrested an Everett man whom police described as part of an “extreme anti-Zionist” organization for allegedly leaving improvised incendiary devices on Boston Common, vandalizing a statue of George Washington, and splattering the State House with paint mixed with feces.
Mixed reviews: High schoolers and educators debated a proposed ban on cellphone use in Massachusetts schools during a Boston City Council hearing yesterday.
Into the breach: An update to Massachusetts law affirmed students’ right to a public education regardless of immigration status or disability, a response to the Trump administration’s efforts to deport immigrants and dismantle the Education Department.
Higher prices: Inflation held steady last month, but Trump’s tariff and immigration policies seem to be raising the cost of some goods and services. (CNBC)
Smithsonian: The White House will vet the museum’s exhibitions and operations ahead of America’s 250th birthday to ensure they conform to Trump’s “directive to celebrate American exceptionalism.” (WSJ 🎁)
Sherrod Brown: The Democratic former US senator from Ohio who lost reelection last year will mount a comeback bid, boosting his party’s chances of retaking the chamber next year. (Axios)
Fraud watch: Federal authorities accused 13 people of operating a call center in the Dominican Republic to scam millions from Americans by posing as grandchildren in distress.
Starstruck: The Vera C. Rubin Observatory opened in northern Chile in June with a mission to survey the southern night sky. The images it has released so far are astounding.
By Teresa Hanafin
💌 Love Letters: After getting duped in a dating scam, this almost-60 letter writer is ready to try again. Meredith’s advice? Get back on apps, join a club, make friends.
🥪 Food as protest: When you’re outraged by the presence of federal agents in D.C., well, sometimes you just have to throw your Subway footlong. At the agents.
🏊 Swimmer’s DBP: Is peeing in the pool really that bad? Do you really want to know? Then again, yes, you do. (HuffPost)
🐶 Dogabunga: The dogs don’t care that it’s a competition. At the North American Diving Dogs event, all they see is sunshine, a welcoming dock, and a gigunda “dog only” swimming pool.
🎭 Funny guy: Liam McGurk has had to take breaks in his comedy career to deal with addiction and mental illness. Now he’s back and making jokes about those very topics.
💪 Guard bod: You stretch before and after workouts. You roll that foam like it’s nobody’s business. Yet that tension in your back or shoulder persists. Why? “Guarding.” (CNN)
🚴 Car-free parents: As electric-assisted bikes improve, parents find it easier than ever to switch to two wheels. For some, minivans are out and e-cargo bikes are in.
Thanks for reading Starting Point.
This newsletter was edited by Annalisa Quinn and Teresa Hanafin.
❓ Have a question for the team? Email us at startingpoint@globe.com.
✍🏼 If someone sent you this newsletter, you can sign up for your own copy.
📬 Delivered Monday through Friday.
Ian Prasad Philbrick can be reached at ian.philbrick@globe.com.